I just don't think it makes sense.

First, 99% of people out there don't care about privacy/cloud. They just want something that works. Therefore, OpenAI/Anthropic/cloud AI lab will soundly beat something like this at a lower cost, smarter, more reliable.

Second, for the 1% who do care, they'll buy Macs with much more memory so they can run much better models instead.

  Javaid offered an example of how Core is meant to work: “Core proactively sends you a notification that you might be getting sick because it sees your resting heart rate is elevated, sees from computer history that you were up late working on a project, and has seasonal context that it is flu season in San Francisco.”

This data almost certainly already lives in the cloud in Apple's server or some smart watch/phone server already. It'd be simple for ChatGPT/Gemini to duplicate something like this over time.

Products like these are the future of AI. Powerful frontier models will become the domain of large corporations and governments and the average person will have their own private fleet of personal AI models for their own uses, occasionally using a hosted model for difficult tasks where privacy is a lesser concern. The prices for AI-capable hardware will fall and then AI will become something that everyone just has in their home or even in their pocket. I believe this because, on both counts, I already do.

I think this is a little ahead of it's time. Until models can be reasonably fast enough on local hardware, I would much rather just pay for one of the major providers.